DIY High performace LD driver

[Benm]

I've been doing some design work on a good current source for laser diodes, for example those from Senkat's group buys. The result of quite some experiments is the shematic attached to this post.

The design in a textbook current source, configured to meet the specific needs of a low-drop, high stability current source. Features:

- Current controllable trough a (low power, plain simple 10k) potentiometer, 0-300 mA range*
- Minimal voltage drop of only 0.9 volts at 200 mA, 1.0 volts at 300 mA.
- Commonly available electronic parts

Some measurements taken driving the red segment of an RGB powerled, Vf=1.99v at 200 mA:

Supply voltage -> output current

2.7 -> 044
2.8 -> 125
2.9 -> 200 (stable from a voltage drop of 2.9 - 1.99 = 0.9 volts)
3.5 -> 202
4.3 -> 199
5.0 -> 200 (calibration point)
10.0 -> 199 **
15.0 -> 202 **

* Range can be controlled by changing Rsense. For example, Rsense of 10 ohms gives a 0-30 mA current source, 0.5 ohms a 0-600 mA version.

** Maximum power of the transistor is 1.25W without a heatsink. For continuous operation at 200 mA, a heatsink is recommended if the input voltage is over 6 volts.

 

[Benm]

This is the schematic above constructed on a breadboard, with components marked. Please excuse the messy setup. Note that i lowerd the current setting to about 25 mA as the LED would otherwise overexpose the photo.

The red led can be any cheap, plain, low-brightness led, this square one was just the first thing i grabbed. Forward drop should be around 1.7V at 10 mA, it's not critical.

 

[chimo]

Very nice.

Looks like you are using a voltage divider based on the forward voltage of the 1n4148. As/If the circuit heats up, the Vf of the 1n4148 should drop a bit and slightly back off the current going to the LD. Have you tried applying a little heat to see how the circuit is affected?

Paul

 

[Benm]

You're correct. I wanted to stick with commonly available parts, so thats where the 1n4148 comes in. There is a predictable drop in current with temperature, going from ambient 25C to a toasty 75C should result in a current reduction of about 10%.

As current drops with rising temperature, this doesn't concern me too much, but this circuit can also work with other references.

 

[Daedal]

This is very nice and quite simple in fact Ben. Thank you for sharing this

--DDL